Rumen escape nitrogen from forages in sheep: comparison of in situ and in vitro techniques using in vivo data

Animal Feed Science and Technology 116 (2004) 35–51

Rumen escape nitrogen from forages in sheep: comparison of in situ and in vitro techniques using in vivo data J.M.J. Gosselink a,b,c,∗ , J.P. Dulphy a , C. Poncet a , J. Aufrère a , S. Tamminga b , J.W. Cone c a

b

INRA, Centre de Clermont-Ferrand-Theix, Unité de Recherches sur les Herbivores, 63122 Saint-Genès, Champanelle, France Wageningen Institute of Animal Sciences, P.O. Box 338, 6700 AH, Wageningen, The Netherlands c Animal Sciences Group of Wageningen UR, Nutrition & Food, P.O. Box 65, 8200 AB, Lelystad, The Netherlands Received 20 August 2003; received in revised form 15 March 2004; accepted 23 April 2004

Abstract The objective of this study was to relate in vivo data of rumen escape N (REN) of forages with REN estimated from models and with determinations of rumen undegradable N. For these determinations and models measurements from in situ and in vitro techniques were used. Eleven forages were investigated in vivo using sheep with cannula in the rumen, duodenum and ileum. These forages were fresh, silage and hay from lucerne and orchard grass, and fresh, silage and haylage from red clover, and silage and hay from perennial ryegrass. Digesta flows were measured with the double marker technique using 51 Cr-EDTA and 103 Ru-phenanthroline. To measure the duodenal flow of microbial nitrogen (N), 15 N was infused as well as purine derivatives were measured in urine excretion. In vivo REN, expressed as g N kg−1 of N intake or as g N kg−1 of duodenal flow of non-ammonia N (NAN), was calculated from duodenal flows of NAN and microbial N and with assumptions for the duodenal flow of endogenous N. REN was also estimated from the models estimating effective undegradable N, using measurements from the in situ nylon bag technique or using Cornell net carbohydrate and protein system with data from CPM (Cornell, Penn, Minor Institute) Dairy Beta program (CPM-REN). With the in situ technique REN was

Abbreviations: ADF, acid detergent fibre; ADIN, ADF insoluble N; BW, body weight; CNCPS, Cornell net carbohydrate and protein system; CP, crude protein; CPM, Cornell, Penn, Minor Institute; CV, coefficient of variation; DM, dry matter; DMI, DM intake; EDN, effective degradable N; EndoN, endogenous N; MN, microbial N; N, nitrogen; NAN, non-ammonia N; NDF, neutral detergent fibre; Nint, N intake; OM, organic matter; PD, purine derivatives; REN, rumen escape N; R.S.E., residual standard error; S.E.M., standard error of the mean ∗ Corresponding author. Tel.: +31 320 237288; fax: +31 320 237230. E-mail address: [email protected] (J.M.J. Gosselink). 0377-8401/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.anifeedsci.2004.04.001

36

J.M.J. Gosselink et al. / Animal Feed Science and Technology 116 (2004) 35–51

calculated from N residues of forages incubated in the rumen, with and without corrections for microbial contamination. These in situ measurements were applied in cows fed a standard diet and in sheep fed the same forage as incubated in the nylon bag. CPM-REN was calculated from five N fractions determined with in vitro techniques. Undegradable N of the 11 forages was measured as N residue after 72 h incubation in nylon bags in the rumen of cows (in situ residual N), after 24 h incubation with protease and as acid detergent insoluble N (ADIN). REN from different in situ measurements and in situ residual N had no relationships with in vivo data. CPM-REN and the in vitro technique using protease had also no relationship with in vivo data. ADIN had a moderate relationship with different in vivo REN determinations and these relations improved when fresh and conserved (silage, hay and haylage) forages were separated (R2 = 0.83–0.87; coefficient of variation = 0.08–0.16). It was concluded, that ADIN has potency to predict in vivo REN of forages. © 2004 Elsevier B.V. All rights reserved. Keywords: Nylon bag; ADIN; CNCPS; Protease; Purine derivatives; 15 N

1. Introduction The duodenal flow of rumen escape nitrogen (N) is an important source of amino acids for ruminants. However, the prediction of this flow is rather difficult, as it depends not only on rumen undegraded protein but also on potentially rumen degradable protein escaping to the duodenum. Models are used to calculate rumen escape nitrogen (REN) from these two nitrogen fractions measured with different techniques. Firstly the in situ technique is used to measure feed N residues after incubation of feed in nylon bags in the rumen and from these residues REN is calculated according to Ørskov and McDonald (1979) or Robinson et al. (1986). Secondly in vitro techniques are used to determine N fractions, which are used to calculate REN, as in the Cornell net carbohydrate and protein system (CNCPS) (Sniffen et al., 1992). In situ and in vitro determinations of REN are based on the hypothesis that rumen undegraded N is a measure for REN. These determinations can be N residue in the nylon bag after 72 h incubation in the rumen, after 24 h incubation with protease (Aufrère and Cartailler, 1988) or as acid detergent insoluble N (Van Soest et al., 1991). The in situ method has been the most widely used method and has commonly been used as reference method (Hvelplund and Weisbjerg, 2000), although in vivo validations of this method are scarce. When using concentrates, Madsen and Hvelplund (1985) observed a close relationship between in vivo and in situ measurements for protein degradation. When using forages Vanzant et al. (1996) observed no significant difference between in vivo and in situ measurements, although in vivo measurements had large standard errors. However, to evaluate in vitro techniques for predicting REN from forages, the in situ technique is not sufficient as long as this method is not well validated. The objective of this study was to relate in vivo REN data of forages with REN of forages determined from models and determinations using in situ and in vitro techniques.

J.M.J. Gosselink et al. / Animal Feed Science and Technology 116 (2004) 35–51

37

2. Materials and methods 2.1. Forages REN of 11 forages was determined in an in vivo experiment. These forages (Table 1) were fresh, silage (with formic acid) and hay from lucerne (Medicago sativa), red clover (Trifolium pratense) and orchard grass (Dactylis glomerata) and silage (with formic acid) and hay from perennial ryegrasses (Lolium perenne). Red clover haylage (a wilted forage wrapped in bales with a dry matter (DM) content of about 500 g/kg forage) was made in stead of red clover hay because of wet harvest circumstances. 2.2. In vivo measurement Six wether sheep, fitted with cannula in the rumen, duodenum and ileum, were used (for orchard grass silage, five wethers). In vivo REN (g/day) was calculated from the duodenal flow of non-ammonia N (NAN, g/day), microbial N (MN, g/day) and endogenous N (EndoN, g/day): (1) in vivo REN = NAN − MN − EndoN The method of the measurement of duodenal digesta flow was as described in Rémond et al. (2003). This method used the double marker technique (Faichney, 1980) with 51 Cr-EDTA and 103 Ru-phenanthroline as flow markers, which were continuously infused into the rumen via separate lines. EndoN was assumed to be 1.5 g N kg−1 of dry matter intake (DMI). This value was based on the assumptions used for sheep fed a diet with only forage, as published in the literature: 1.5 g NAN per day (Siddons et al., 1979), 2.0 g N kg−1 of DM intake per day (Beever et al., 1987), 2.5 g NAN per day (Kawas et al., 1990). Table 1 Composition (g kg−1 of dry matter) of the 11 forages Forage

Method of conservation

Dry matter

Ash

Crude protein

NDFa

ADFb

Lucerne

Fresh Silage Hay

162 212 861

138 98 99

198 182 171

498 438 560

346 328 379

Red clover

Fresh Silage Haylage

127 171 524

120 92 108

168 166 128

492 478 475

348 343 352

Orchard grass

Fresh Silage Hay

193 217 852

80 71 70

116 126 110

676 614 697

360 343 376

Lolium perenne

Silage Hay

191 873

92 96

101 91

578 632

371 382

a b

NDF: neutral detergent fibre. ADF: acid detergent fibre.

38

J.M.J. Gosselink et al. / Animal Feed Science and Technology 116 (2004) 35–51

Table 2 The amount of microbial purines absorbed fron the intestines (mmol/day) estimated from the individual purine derivatives (allantoin, uric acid, xanthine and hypoxanthine), which were excreted in the urine of sheep fed the 11 forages of Table 1 Method of conservation

Allantoin

Lucerne

Fresh Silage Hay

18.88 18.66 12.62

3.11 1.57 1.21

1.64 1.87 1.53

Red clover

Fresh Silage Haylage

16.76 12.32 12.69

1.46 1.13 1.05

2.53 1.27 1.15

Orchard grass

Fresh Silage Hay

14.90 13.95 11.41

1.90 1.50 1.00

1.12 1.90 1.45

Lolium perenne

Silage Hay

13.46 13.76

1.73 1.47

1.61 1.17

65 14.9 0.35

65 1.59 0.101

65 1.60 0.065

n Mean S.E.M.

Uric acid

Xanthine + hypoxanthine

Forage

Two methods were applied to measure MN. In the first method, the microbes in the rumen were marked with 15 N enriched (>0.98) ammonium sulphate, which was continuously infused into the rumen in the same period as the flow markers (Rémond et al., 2003). In the second method (Chen and Gomes, 1992a), the urine excretion of purine derivatives (PD: xanthine, hypoxanthine, uric acid and allantoin) from the six sheep were measured. They were analysed with a spectrophotometer: xanthine plus hypoxanthine (Chen et al., 1990a), uric acid and allantoin (Fujihara et al., 1987) (Table 2). The equation (Chen et al., 1990b) used for sheep to describe the quantitative relationship between absorption of microbial purines from the intestines (X: mmol/day) and excretion of PD in urine (Y: mmol/day) is: Y = 0.84X + (0.150W 0.75 × e−0.25X )

(1)

where W is the body weight of the sheep (between 45 and 60 kg). From X the intestinal flow of microbial N (g N/day) is estimated: MN =

70X = 0.727X 0.83 × 0.116 × 1000

(2)

where 0.83 is the digestibility of microbial purines; 70 is the N content of purines (mg N/ mmol); 0.116 is the ratio of purine N to total N in mixed rumen microbes (Chen et al., 1992b). These two MN measurements resulted in two estimations of in vivo REN, expressed as g N kg−1 of N intake: REN-15 N and REN-PD. When in vivo REN is expressed as part of duodenal NAN flow, the contributions of this N source to the production of amino aids absorbed in the small intestine can be calculated. Therefore, two other estimations of in vivo REN were used, expressed as g N kg−1 of duodenal NAN flow: RENAN-15 N and RENAN-PD.

J.M.J. Gosselink et al. / Animal Feed Science and Technology 116 (2004) 35–51

39

2.3. In situ measurement The 11 forages were incubated in nylon bags in the rumen of cows and sheep. The cows were fed a diet of hay (0.7 of DMI: dry matter content of 850 g/kg forage, crude protein (CP) content of 106 g/kg DM and neutral detergent fibre (NDF) content of 652 g/kg DM) and concentrate (0.3 of DMI: CP content of 140 g/kg DM and NDF content of 290 g/kg DM). The sheep were fed a diet with only the forage, which was also incubated in the nylon bags, without concentrate. The method of sample preparation was described by Dulphy et al. (1999). Summarised briefly, the samples of fresh forages and silages were lacerated to a particle size of 4–5 mm, put into nylon bags, then quickly frozen in liquid N, preserved at −20 ◦ C for some months and thawed slowly and incubated in the rumen. Rapid freezing of fresh forage in liquid nitrogen had no effect on the ruminal degradation kinetics compared with unfrozen fresh forage, although results in the literature concerning freezing are conflicting (Dulphy et al., 1999). Hays were only ground to a mesh size of 4 mm and put into the nylon bags (Ankom, pore size 30–60 ␮m, internal surface of 5 cm × 11 cm and closed by two stitches). The in situ measurements were performed according to the procedures described by Michalet-Doreau et al. (1987). The results from in situ measurements of the 11 forages in sheep were taken from Aufrère et al. (2000, 2002, 2003). These measurements had incubation periods of 2, 4, 8, 16, 24 and 48 h and used four sheep. Because of the rumen capacity only two replications per sheep were used for 2, 4 and 8 h, whereas three replications were used for 16, 24 and 48 h to sample sufficient residues for chemical analyses. Using this method the soluble fraction was determined by soaking the bags with the forage in warm water (40 ◦ C) during 1.5 h followed by drying, in contrast with the in situ method in cows which used the soluble fraction calculated by the model (Ørskov and McDonald, 1979), which is explained later. In this latter method, six samples per forage were incubated in the cow rumen for each incubation period (0, 2, 4, 8, 16, 24, 48 or 72 h of incubation). Three rumen fistulated cows were used and each forage was incubated in duplicate at two different days per cow. After incubation the bags were kept at −20 ◦ C until analysis. Prior to the analysis, the bags were thawed and then rinsed with cold water until the water ran clear. In both methods, the bags were subjected to vigorous mechanical pummelling between two metal plates for 7 min in a Colworth Stomacher (Merry and Mc Allan, 1983), followed by further washing to remove bacteria and finally dried at 60 ◦ C for 48 h. Michalet-Doreau and Ould-Bah (1989) showed that pummelling or beating the residues in a stomacher could significantly reduce microbial contamination of the undegraded fraction of the sample. DM content of the six residues from the nylon bags was determined and then the residues were pooled per incubation time for N analyses. The N residues were fitted using the first order model of Ørskov and McDonald (1979). From this model the degradation and the degradation rate of N in the rumen are calculated: Disappearance(fraction of N incubated) at t = a + b(1 − exp−ct )

(3)

40

J.M.J. Gosselink et al. / Animal Feed Science and Technology 116 (2004) 35–51

where a is the soluble N fraction; b the degradable fraction of N; c the degradation rate per hour; = the incubation time. From these parameters effective degradable N (EDN) was calculated using a fractional passage rate (kp) of 0.06 h−1 (Michalet-Doreau et al., 1987):
 bc EDN(fraction of N incubated) = a + (4) c + kp EDN was also calculated from N residues in the nylon bags, which were corrected for residual microbial contamination of the undegraded fraction in the nylon bags according to the equation of Ould-Bah (1989):   N in residue corrected N initial incubated     N in residue DM of residue = not corrected − z × 6.67 (5) N initial incubated N initial incubated where z is the DM of bacteria in residue/DM of residue; z = 1 (after 2 h of in situ incubation (incub)), 1.3 (after 4 h incub), 1.9 (after 8 h incub), 3.1 (after 16 h incub) and 4 (after 24, 48 and 72 h incub); 6.67 = average content of bacterial N in bacterial DM. In situ REN was calculated as effective undegraded N, which is 1000 minus EDN (fraction of N incubated) and compared with in vivo REN. The N residues (g N g−1 of N initial incubated, corrected for microbial contamination as in Eq. (5)) in the nylon bags after 72 h of incubations in the rumen of the cows (in situ residual N) were also compared with in vivo REN. 2.4. In vitro measurement In vitro techniques were used to measure N fractions. Five N fractions were measured to estimate REN from the model of Cornell net carbohydrate and protein system (CNCPS) as described by Sniffen et al. (1992). In vivo REN was also related to acid detergent insoluble nitrogen (ADIN) determined as described by Van Soest et al. (1991) and to indigestible N after incubation with protease (protease N: Aufrère and Cartailler, 1988). The model of CNCPS is nowadays implemented in the CPM (Cornell, Penn, Minor Institute) Dairy Beta program (CPM-Dairy, 2003). To calculate REN from this model (CPM-REN) crude protein was partitioned into five fractions, which were analysed as described by Sniffen et al. (1992). Fractions A and B1 are soluble in buffer (Krishnamoorthy et al., 1983). Fraction A is non-protein N, which is soluble in phosphate-borate buffer (pH 6.7) and in trichloracetic acid (TCA). Fraction B1 is rapidly degradable true protein and is TCA-precipitated protein from the buffer-soluble fraction. Fraction C is unavailable protein bound to cell walls and is derived from acid detergent insoluble nitrogen (ADIN, fraction of N total) (Van Soest et al., 1991). Fraction B3 is slowly degradable protein and is neutral detergent insoluble nitrogen (NDIN, fraction of N total) (Van Soest et al., 1991) minus ADIN. Fraction B2 is the remaining N and is true protein with an intermediate degradation rate between fraction B1 and B3. Degradation and passage rates were obtained from version 2.0.25a of CPM-Dairy Beta program (CPM-Dairy, 2003).

J.M.J. Gosselink et al. / Animal Feed Science and Technology 116 (2004) 35–51

41

ADIN data were taken from the data measured for CPM-REN (Van Soest et al., 1991). Indigestible N was measured after 24 h of incubation with proteases in borate/ phosphate buffer at pH 8.0 as described by Aufrère and Cartailler (1988) and Cone et al. (1996). The protease from S. griseus (typeXIV, Sigma P-5147, St Louis, MO, USA) was used in a concentration of 20 mg/l buffer. Tetracyclin (1 mg/l, Sigma N-3503) and nystatin (10 mg/l, Sigma T-3258) were added to the buffer to prevent microbial growth. 2.5. Chemical analysis DM contents of feed and residues in nylon bags were determined by drying at 80 ◦ C for 48 h and ash content was determined after 6 h at 550 ◦ C. The DM of silage and haylage was corrected for fermentation products (Dulphy et al., 1975). N was determined using the Kjehldahl method (AOAC, 1980), except for N in the residues in the nylon bags, which was determined with the method of Dumas (Merz, 1968). NDF and acid detergent fibre (ADF) were determined on the samples dried at 60 ◦ C using the method described by Van Soest et al. (1991) and were expressed with residual ash. NDF was determined without the use of sodium sulphite and alpha amylase. 2.6. Statistics Statistical analyses were done with procedures of GenStat (2002).The datasets were described with the mean and standard error of the mean (S.E.M.). To relate REN determined from models and determinations using in situ and in vitro techniques with in vivo REN data, the following model was used to: in vivo REN = β0 + β1 × methodj + factor l + εjl

(6)

where methodj is the REN from the model using the in situ technique (j = 1), CPM-REN (j = 2), in situ residual N (j = 3), ADIN (j = 4) or protease N (j = 5); factorl the forage family (legume or grass) (l = 1), method of conservation (fresh or conserved) (l = 2); β0–1 the regression coefficients; εjl the residual error, supposed to be normal distributed with zero mean and constant (residual) standard error (R.S.E.). P value of the equations and estimates were considered significant when lower than 0.1. Relationships were described with R2 , R.S.E. and coefficient of variation (CV), which is the ratio of R.S.E. and the mean of the observed data.

3. Results 3.1. Forages The lucerne forages, fresh red clover and red clover silage had a high CP content compared to the other forages (Table 1) and their ratio of duodenal NAN flow and N intake was lower than 1 (Table 3). Red clover haylage had the highest REN and fresh orchard grass had the lowest REN.

42

J.M.J. Gosselink et al. / Animal Feed Science and Technology 116 (2004) 35–51

Table 3 In vivo data of 11 forages: DM intake (g/day), the ratio of duodenal non-ammonia nitrogen (NAN) flow and nitrogen (N) intake (NAN/Nint) and rumen escape N (REN: fraction of N intake; RENAN: fraction of duodenal NAN flow) measured in vivo using 15 N (REN-15 N, RENAN-15 N) or purine derivatives (REN-PD, RENAN-PD) REN-15 N

RENAN-15 N

Forage

Method of conservation

DM intake

NAN/ Nint

Lucerne

Fresh Silage Hay

1528 1686 1166

0.85 0.83 0.89

0.20 0.24 0.27

0.24 0.29 0.30

0.40 0.39 0.42

0.47 0.47 0.48

Red clover

Fresh Silage Haylage

1287 1335 1290

0.91 0.91 1.18

0.11 0.29 0.38

0.13 0.21 0.32

0.33 0.50 0.64

0.36 0.54 0.53

Orchard grass

Fresh Silage Hay

1332 1320 1161

1.09 1.06 1.18

0.06 0.20 0.24

0.05 0.19 0.21

0.38 0.43 0.51

0.35 0.40 0.44

Lolium perenne

Silage Hay

1271 1300

1.37 1.43

0.30 0.21

0.22 0.15

0.58 0.58

0.42 0.40

65 1.07 0.026

65 0.23 0.011

65 0.22 0.011

65 0.47 0.014

65 0.44 0.009

n Mean S.E.M.

65 1334 20.0

REN-PD

RENAN-PD

3.2. In vivo data In vivo REN values calculated from in vivo data using 15 N as microbial marker were lower than REN values calculated from in vivo data using PD for the measurement of microbial N synthesis in the rumen (Table 3). However, these measurements of microbial N (g/day) using 15 N and PD were related: microbial N from PD = 0.69 × microbial N from 15 N (n = 11; R2 = 0.94; R.S.E. = 0.67; CV = 0.043). The difference between REN per kg N intake (REN-15 N or REN-PD) and REN per kg of duodenal NAN flow (RENAN-15 N or RENAN-PD) was the result of an unstable variable ratio of duodenal NAN flow and N intake (Table 3). The relationships between in vivo REN and in situ or in vitro REN did not improve using the different assumptions for the duodenal flow of endogenous N, as mentioned in the material and methods, to calculate the in vivo REN. Other assumptions, 0.181 g endogenous N per kg body weight (BW)0.75 (Ørskov et al., 1986) or 0.279 g endogenous N per kg BW0.75 (Lintzenich et al., 1995), resulted in similar relationships between in vivo REN and in situ or in vitro REN as when using 1.5 g endogenous N per kg of DM intake. 3.3. In situ technique Procedures to calculate EDN differed slightly between cows and sheep (Table 4) and the results obtained with these procedures were poorly related (R2 was between 0.36 and 0.43). EDN values differed between cows and sheep, although this difference was small when concerning lucerne and large when concerning fresh red clover. The trends in EDN followed the trends in soluble N fraction. The results suggest that the forages had a higher

Forage

Method of conservation

Cow plus a

Cow min b

kd

EDN

a

Sheep plus

b

kd

EDN

a

b

Sheep min kd

EDN

a

b

kd

EDN

Lucerne

Fresh Silage HAY

0.48 0.67 0.33

0.42 0.23 0.55

0.19 0.13 0.11

0.79 0.82 0.69

0.47 0.66 0.33

0.39 0.21 0.53

0.18 0.12 0.10

0.77 0.80 0.66

0.41 0.60 0.34

0.45 0.29 0.54

0.30 0.22 0.12

0.78 0.83 0.69

0.41 0.60 0.35

0.42 0.26 0.52

0.30 0.21 0.10

0.76 0.80 0.66

Red clover

Fresh Silage Haylage

0.66 0.49 0.59

0.30 0.46 0.36

0.17 0.17 0.08

0.89 0.83 0.80

0.66 0.49 0.59

0.28 0.44 0.34

0.17 0.17 0.07

0.87 0.81 0.77

0.25 0.42 0.41

0.55 0.45 0.51

0.35 0.22 0.09

0.71 0.76 0.70

0.25 0.42 0.41

0.53 0.41 0.47

0.35 0.22 0.08

0.69 0.74 0.66

Orchard grass

Fresh Silage Hay

0.33 0.65 0.38

0.61 0.28 0.52

0.10 0.12 0.08

0.71 0.84 0.67

0.32 0.65 0.37

0.58 0.26 0.49

0.09 0.09 0.06

0.67 0.81 0.61

0.36 0.53 0.33

0.57 0.40 0.60

0.11 0.08 0.05

0.72 0.75 0.60

0.36 0.52 0.32

0.51 0.37 0.61

0.10 0.06 0.04

0.67 0.71 0.54

Lolium perenne Silage Hay

0.69 0.31

0.25 0.62

0.19 0.12

0.89 0.72

0.69 0.31

0.21 0.56

0.15 0.10

0.84 0.67

0.57 0.17

0.31 0.73

0.20 0.08

0.81 0.60

0.57 0.17

0.26 0.69

0.19 0.07

0.76 0.54

n Mean S.E.M.

11 11 11 11 11 11 11 11 11 11 11 11 11 11 11 11 0.51 0.42 0.13 0.79 0.51 0.39 0.12 0.75 0.40 0.49 0.17 0.72 0.40 0.46 0.16 0.68 0.005.8 0.043 0.013 0.024 0.046 0.042 0.013 0.026 0.039 0.039 0.030 0.023 0.040 0.040 0.032 0.026

J.M.J. Gosselink et al. / Animal Feed Science and Technology 116 (2004) 35–51

Table 4 Rumen effective degradable nitrogen (EDN: fraction of N incubated) of the 11 forages of Table 1, calculated according to Ørskov and McDonald (1979) with a passage rate of 0.06 h−1 and from soluble N fraction (a: fraction of N incubated), degradable N fraction (b: fraction of N incubated) and the degradation rate of N (kd, fraction·h−1 ), measured in cows (standard diet) or in sheep (same diet as in nylon bag), with (plus) and without (min) the correction of Ould-Bah (1989)

43

44

Forage

Method of conservation

CPM-REN

A

B1

B2 kd

B3 kd

C

kp

Protease N

kd

In situ residual N

Lucerne

Fresh Silage Hay

0.21 0.15 0.26

0.31 0.57 0.28

0.028 0.021 0.034

2.00 1.50 1.50

0.52 0.32 0.57

0.15 0.11 0.09

0.097 0.059 0.065

0.020 0.018 0.01325

0.054 0.037 0.054

0.076 0.070 0.063

0.40 0.23 0.46

0.17 0.29 0.18

Red clover

Fresh Silage Haylage

0.15 0.19 0.19

0.42 0.40 0.51

0.012 0.018 0.005

2.00 1.50 1.50

0.48 0.47 0.37

0.15 0.11 0.09

0.053 0.066 0.057

0.020 0.018 0.013

0.035 0.042 0.064

0.076 0.070 0.063

0.24 0.26 0.18

0.10 0.10 0.14

Orchard grass

Fresh Silage Hay

0.25 0.21 0.28

0.42 0.45 0.36

0.068 0.020 0.022

2.00 2.00 1.50

0.32 0.40 0.44

0.11 0.09 0.11

0.173 0.119 0.162

0.020 0.018 0.015

0.019 0.013 0.020

0.066 0.054 0.055

0.22 0.26 0.30

0.10 0.14 0.17

Lolium perenne

Silage Hay

0.17 0.26

0.46 0.17

0.014 0.006

2.00 1.50

0.43 0.71

0.09 0.11

0.065 0.089

0.018 0.015

0.030 0.023

0.061 0.064

0.23 0.36

0.19 0.12

11 0.28 0.013

22 0.40 0.023

22 0.023 0.0042

22 0.29 0.018

11 0.14 0.076

n Mean S.E.M.

22 0.46 0.023

22 0.091 0.0089

22 0.035 0.0035

J.M.J. Gosselink et al. / Animal Feed Science and Technology 116 (2004) 35–51

Table 5 Rumen escape nitrogen (REN: fraction of N intake) calculated from rumen passage rates (kp, fraction h−1 ), and N fractions [A, B1, B2, B3, C (acid detergent insoluble N): fraction of forage N] and their degradation rates (kd, fraction h−1 ), as described in the CPM (Cornell, Penn, Minor Institute) system (CPM-REN), and undegradable N (fraction of forage N) measured in vitro with proteases (Protease N) and in situ as the N residue in nylon bags after 72 h of incubation in the rumen (in situ residual N), using the forages of Table 1

J.M.J. Gosselink et al. / Animal Feed Science and Technology 116 (2004) 35–51

45

Table 6 Relationships (R2 , residual standard error (R.S.E.), coefficient of variation (CV) and n = 11) between in vivo rumen escape nitrogen (REN-15 N, RENAN-15 N, REN-PD, RENAN-PD) and rumen escape nitrogen (REN) determined using in situ measurements in cows and sheep (situ-cow-REN, situ-sheep-REN) and REN using in vitro measurements as in the CPM (Cornell, Penn, Minor Institute) system (CPM-REN) and the determinations of acid detergent insoluble N (ADIN), protease N and in situ residual N REN-determination

REN-15 N

RENAN-15 N

REN-PD

RENAN-PD

Situ-cow-REN Situ-sheep-REN CPM-REN ADIN

P P P P R2 R.S.E. CV

NSa NSa NSa <0.1 0.25 75.9 0.33

NSa NSa NSa <0.05 0.46 62.2 0.28

NSa NSa NSa NSa

NSa NSa NSa <0.05 0.47 51.2 0.12

Protease N In situ residual N

P P

NSa NSa

NSa NSa

NSa NSa

NSa NSa

a

NS: non-significant (P > 0.1).

degradable N fraction (b in Table 4) and degradation rates of N (c in Table 4) in sheep than in cows. The correction of Michalet-Doreau and Ould-Bah (1989) resulted in a small increase of the degradable N fraction, the degradation rates and EDN. The four EDN measurements were used to estimate in situ REN or effective undegradable N. The values of in situ REN measured in cows and sheep had no significant relationships with in vivo REN (Table 6). These relationships did not improve when fractional passage rates other than 0.06 h−1 were used to calculate EDN. Also in situ residual N (Table 5) showed no significant relationships with in vivo data (Table 6). 3.4. In vitro techniques CPM-REN calculated from five N fractions (Table 5) was not related with in vivo REN (Table 6) and poorly related with in situ REN (Table 7), measured with cows or sheep on a standard ration and corrected for microbial contamination according to Ould-Bah Table 7 Relationships (R2 , P < 0.05) between rumen escape nitrogen N (REN) determined using in situ measurements in cows and sheep (situ-cow-REN, situ-sheep-REN) and using in vitro measurements as in the CPM (Cornell, Penn, Minor Institute) system (CPM-REN) and the determinations of acid detergent insoluble N (ADIN), protease N and in situ residual N REN-determination

Situ-cow-REN

Situ-cow-REN Situ-sheep-REN CPM-REN ADIN Protease In situ residual N

NSa

a

0.35 0.40 NSa NSa NSa

NS: non-significant (P > 0.05).

Situ-sheep-REN

CPM-REN

ADIN

Protease

NSa 0.40 NSa NSa NSa

NSa 0.58 NSa

NSa NSa

NSa

46

J.M.J. Gosselink et al. / Animal Feed Science and Technology 116 (2004) 35–51

(1989). Nevertheless the values from in situ REN measured with sheep (situ-sheep-REN) and CPM-REN were close: situ-sheep-REN = 0.997 * CPM-REN (Table 7: P < 0.001; R2 = 0.40; R.S.E. = 5.90 and CV = 0.212). CPM-REN had a moderate relationship with indigestible N measured using incubation with protease (Tables 5 and 7). Of all determinations ADIN was best related with in vivo REN (Table 6), but was not related with other determinations (Table 7). The relationship between ADIN and RENAN-PD had a lower CV than the relationship between ADIN and REN-15 N or RENAN-15 N (Table 6). These relationships improved when the method of conservation (fresh or conserved: silage, haylage and hay) was included in the regression analysis: (1) REN-15 N = 3.08 × ADIN + 1.6 (fresh forage) 15.7 (conserved forage); (2) RENAN-15 N = 3.72 × ADIN + 0.7 (fresh) 11.6 (conserved); (3) RENAN-PD = 2.74 × ADIN + 29.4 (fresh) 36.2 (conserved);

R2 = 0.87, CV = 0.14 R2 = 0.83, CV = 0.16 R2 = 0.83, CV = 0.08

In these regressions ADIN and method of conservations are orthogonal.

4. Discussion 4.1. In vivo data In vivo REN differed between the calculations from 15 N and the calculations from PD, due to their fractional difference of 0.31 in the measurement of microbial N synthesis in the rumen. This value of 0.31 is close to the percentages observed by Perez et al. (1996). This difference between these microbial N measurements can partly be explained by the contamination of endogenous protein with 15 N, but mainly be explained by the determination of urinary excreted PD. Using the Eqs. (1) and (2) (Section 2), a dilution factor of urine samples for the determination of urinary excreted PD was found, when the reference value for microbial N synthesis was 32 g microbial N kg−1 digestible organic matter (OM) fermented in the rumen (Chen and Gomes, 1992a). This value is 26% lower than the mean value of 43.2 g microbial N kg−1 OM apparently digested in the rumen (n = 11 forages) observed with 15 N as microbial marker in the in vivo experiment used for this study. Nevertheless measurements of urine excretion of PD proved useful in many studies on microbial protein production from forage feeding (Tamminga and Chen, 2000). In this study, the comparisons with in vivo REN as part of the duodenal NAN flow were also presented as this ratio varied less than the ratio in vivo REN as part of N intake, as used commonly. Duodenal NAN flow does not include forage N degraded to ammonia disappearing from the rumen and thus not contributing to escape N and microbial N. When forages had a high CP content, the ratio of duodenal NAN flow and N intake was lower than 1 (Tables 1 and 3), meaning a high ammonia production in the rumen. This ratio was higher than 1 when forages had a low CP (<160 g kg−1 DM) (Tables 1 and 3), as a result

J.M.J. Gosselink et al. / Animal Feed Science and Technology 116 (2004) 35–51

47

of the utilisation of N from the urea recycling in the sheep. When escape N or microbial N were expressed as part of duodenal NAN flow, the contributions of these two N sources to the production of amino acids absorbed in the small intestine can be calculated. 4.2. In situ technique Effective degradable N differed between the in situ measurements in cows and the measurements in sheep, as a result of different methods of determining soluble N. Determining the soluble N fraction by rinsing and soaking or washing in warm water is more commonly used than the calculation from the model. The difference in degradation rate of N between the measurements in sheep and cows could be due to a combination of species and ration fed during the incubations. Comparing cows and sheep (Texel ewes), cows displayed 40% lower degradation rates, whereas dietary roughage: concentrate ratio had no effect (Šebek and Everts, 1999). Cows fed the same forage as incubated with the nylon bags, showed higher degradation rates than cows fed a “standard” brome hay (Vanzant et al., 1996). Despite all the different in situ procedures and calculations, not even a slight relationship between in situ and in vivo results was found. Compared to in vivo, EDN as calculated in protein evaluation systems (Vérité et al., 1987; Tamminga et al., 1994) is less dynamic. These systems used three N fractions, a fixed passage rate and the soluble fraction assumed to be totally degraded. Recently, it was proven that part of the soluble N escapes rumen fermentation and enters the duodenum as amino acids (Aufrère et al., 2000, 2002, 2003). The in situ technique is generally accepted to obtain REN values for feed protein evaluation systems and as reference (Hvelplund and Weisbjerg, 2000). Moreover, a tendency towards more confidence in in situ measurements than in vivo measurements is found in literature. But validations of in situ rumen escape N from forages with in vivo data are scarce, whereas other recent simple and extensive rumen models already have been validated with in vivo data (Bannink et al., 1997; Bateman et al., 2001). As long as both measurements have important limitations, which have often been reviewed (Hvelplund and Weisbjerg, 2000; Firkins et al., 1998), this study used in vivo data as most reliable reference values. The most important limitations of in situ measurements are its low repeatability and its lack of reproducibility according to Michalet-Doreau and Ould-Bah (1992) and Hvelplund and Weisbjerg (2000), although in this study the same conclusion on repeatability could not be made in the procedures with sheep and cows because of pooled samples per incubation time. However, in vivo data can also have high standard errors. Not much is known about repeatability and reproducibility of in vitro methods. Vanzant et al. (1996) observed standard errors from the in vivo techniques nearly five times larger than the standard errors from the in situ techniques. 4.3. In vitro techniques Calsamiglia et al. (2000) reviewed the in vitro techniques to predict protein degradation and they concluded that for improving these techniques, further understanding of protein degradation and utilization by rumen microbes and the ruminant animal is necessary.

48

J.M.J. Gosselink et al. / Animal Feed Science and Technology 116 (2004) 35–51

However, interesting results with forages were observed in this study concerning the determination of REN or undegraded N by in vitro techniques. CPM-REN data were closer to the in situ REN data from sheep than from cows, probably due to the determination of the soluble fraction. Compared to the in situ technique, the determination of CPM-REN data appeared to be further away from copying the real rumen process. Nevertheless CPM-data will have higher repeatability and reproducibility than in situ measurements, because of chemical and in vitro determinations of the N fractions and moreover the degradation rates and passage rates were assumed and obtained from CPM-Dairy (2003). Some authors suggest that Cornell net carbohydrate and protein system predicts REN or N fractions passing to the duodenum better than NRC (1989), which is based on in situ measurements (Van Amburgh et al., 1998; Bateman et al., 2001). Most investigations using proteases from Streptomyces griseus are done with concentrates (Calsamiglia et al., 2000). Aufrère et al. (1989) investigated forages and concluded that protein degradation of hay could be predicted with CP content and that protease improved the relationship for heat dried hays. In our study, protease plus CP content had no relationship with in vivo REN and a moderate relationship with REN determined from in situ measurements (R2 < 0.50). ADIN can be a simple measurement for predicting REN of forages. ADIN is N associated with lignin as well as indigestible nitrogen (Thomas et al., 1982) or unavailable nitrogen (Van Soest, 1982). ADIN was also related with undegradable N from in situ measurement with forages (Vanzant et al., 1996), in contrast to the results in this study. Generally ADIN is used as a measure for heat damage and is used as indigestible or unavailable N in rumen N models (Van Soest, 1982; Sniffen et al., 1992). Modest heating increases ADIN as well as N escaping rumen fermentation and may result in an increased supply of N absorbed from the small intestine (Merchen and Bourquin, 1994; Yang et al., 1993). Conserved forages may receive heat during drying, wilting and stocking and consequently their ADIN fraction may show a different relationship with in vivo REN compared to fresh forages. The part of REN, which is potentially rumen degradable N, entering the duodenum is more difficult to explain by ADIN than the part of REN, which is rumen undegradable N. However, ADIN was related with apparent N digestibility (Thomas et al., 1982) and with forage digestibility (Van Soest, 1982). Increasing ADIN decreases digestibility and consequently rumen passage rate, which affects REN.

5. Conclusion ADIN has potency to predict in vivo REN, although this prediction needs validation. The determination of ADIN is cheap, fast and does not impair animal welfare.

Acknowledgements The authors are very grateful for the assistance from Unité de Recherches sur les Herbivores (INRA, Centre de Clermont-Ferrand-Theix, France) and Animal Sciences Group of Wageningen UR (Nutrition & Food, Lelystad, The Netherlands).

J.M.J. Gosselink et al. / Animal Feed Science and Technology 116 (2004) 35–51

49

References AOAC, 1980. Official Methods of Analysis, 13th ed. Association of Official Analytical Chemists. Washington, DC. Aufrère, J., Cartailler, D., 1988. Mise au point d’une méthode de laboratoire de prévision de la dégradabilité des protéines alimentaires des aliments concentrés dans le rumen. Ann. Zootech. 37, 255–270. Aufrère, J., Graviou, D., Baumont, R., Detour, A., Demarquilly, C., 2000. Degradation in the rumen of proteins from fresh lucerne forage in various stages of growth and conserved as silage or hay. Ann. Zootech. 49, 461–474. Aufrère, J., Graviou, D., Demarquilly, C., 2002. Protein degradation in the rumen of red clover forage at various stages of growth and conserved as silage or wrapped big bales. Reprod. Nutr. Dev. 42, 559–572. Aufrère, J., Graviou, D., Demarquilly, C., 2003. Ruminal degradation of protein of cocksfoot and perennial ryegrass as affected by various stages of growth and conservation methods. Anim. Res. 52, 245–261. Aufrère, J., Michalet-Doreau, B., Graviou, D., Vérité, R., 1989. Predicting in sacco degradability of hay protein by chemical or enzymatic methods. In: Proceedings of the XVIth International Grassland Congress. Nice, France, pp. 887–888. Bannink, A., De Visser, H., Van Vuuren, A.M., 1997. Comparison and evaluation of mechanistic rumen models. Br. J. Nutr. 78, 563–581. Bateman, H.G., Clark II, J.H., Patton, R.A., Peel, C.J., Schwab, C.G., 2001. Prediction of crude protein and amino acid passage to the duodenum of lactating cows by models compared with in vivo data. J. Dairy Sci. 84, 665–679. Beever, D.E., Losada, H.R., Gale, D.L., Spooner, M.C., Dhanoa, M.S., 1987. The use of monensin or formaldehyde to control the digestion of the nitrogenous constituents of perennial ryegrass (Lolium perenne cv. Melle) and white clover (Trifolium repens cv. Blanca) in the rumen of cattle. Br. J. Nutr. 57, 57–67. Calsamiglia, S., Stern, M.D., Bach, A., 2000. Enzymatic and microbial-cell preparation techniques for predicting rumen degradation and postruminal availability of protein. In: Givens, D.I., Owen, E., Axford, R.F.E., Omed, H.M. (Eds.), Forage Evaluation in Ruminant Nutrition. CABI Publishing, Wallingford, UK, pp. 259–279. Chen, X.B., Mathieson, J., Hovell, F.D.D., Reeds, P.J., 1990a. Measurement of purine derivatives in urine of ruminants using automated methods. J. Sci. Food Agric. 53, 23–33. Chen, X.B., Hovell, F.D.D., Ørskov, E.R., Brown, D.S., 1990b. Excretion of purine derivatives by ruminants: effect of exogenous nucleic acid supply on purine derivative excretion by sheep. Br. J. Nutr. 63, 131–142. Chen, X.B., Gomes, M.J., 1992a. Estimation of microbial protein supply to sheep and cattle based on urinary excretion of purine derivatives—an overview of the technical details. Occasional Publication of the International Feed resources Unit. Rowett Research Institute, Aberdeen, UK. Chen, X.B., Chen, Y.K., Franklin, M.F., Ørskov, E.R., Shand, W.J., 1992b. The effect of feed intake and body weight on purine derivative excretion and microbial protein supply in sheep. J. Anim. Sci. 70, 1534–1542. Cone, J.W., Van Gelder, A.H., Steg, A., Van Vuuren, A.M., 1996. Prediction of in situ rumen escape protein from in vitro incubation with proteases from Streptomyces griseus. J. Sci. Food Agric. 72, 120–126. CPM-Dairy Beta Program, 2003. Cornell∗ Penn∗ Miner—Dairy, version 2.0.25a. School of Veterinary Medicine, University of Pennsylvania. Available: http://www.mail.vet.upenn.edu/, May 2003. Dulphy, J.P., Demarquilly, C., Baumont, R., Jailler, M., L’Hotelier, L., Dragomir, C., 1999. Study of modes of preparation of fresh and conserved forage samples for measurement of their dry matter and nitrogen degradations in the rumen. Ann. Zootech. 48, 275–288. Dulphy, J.P., Demarquilly, C., Henry, M., 1975. Perte de composés volatils lors de la détermination à l’étuve de la teneur en matière sèche des ensilages. Ann. Zootechn. 24, 743–756. Faichney, G.J., 1980. The use of markers to measure digesta flow from stomach of sheep fed once daily. J. Agric. Sci. (Camb.) 94, 313–318. Firkins, J.L., Allen, M.S., Oldick, B.S., St-Pierre, N.R., 1998. Modelling ruminal digestibility of carbohydrates and microbial protein flow to the duodenum. J. Dairy Sci. 81, 3350–3369. Fujihara, T., Ørskov, E.R., Reeds, P.J., 1987. The effect of protein infusion on urinary excretion of purine derivatives in ruminants nourished by intragastric nutrition. J. Agric. Sci. Camb. 109, 7–12. GenStat. 2002. Reference manual, Release 6.1. VSN International, Oxford, UK. Hvelplund, T., Weisbjerg, M.R., 2000. In situ techniques for the estimation of protein degradability and postrumen availability. In: Givens, D.I., Owen, E., Axford, R.F.E., Omed, H.M. (Eds.), Forage Evaluation in Ruminant Nutrition. CABI Publishing, Wallingford, UK, pp. 233–258.

50

J.M.J. Gosselink et al. / Animal Feed Science and Technology 116 (2004) 35–51

Kawas, J.R., Jorgensen, N.A., Lu, C.D., 1990. Influence of alfalfa maturity on feed intake and site of nutrient digestion in sheep. J. Anim. Sci. 68, 4376–4386. Krishnamoorthy, U.C., Sniffen, C.J., Stern, M.D., Van Soest, P.J., 1983. Evaluation of a mathematical model of rumen digestion and an in vitro simulation of rumen proteolysis to estimate the rumen-undegraded nitrogen content of feedstuffs. Br. J. Nutr. 50, 555–568. Lintzenich, B.A., Vanzant, E.S., Cochran, R.C., Beaty, J.L., Brandt Jr., R.T., St. Jean, G., 1995. Influence of processing supplemental alfalfa on intake and digestion of dormant bluestem-range forage by steers. J. Anim. Sci. 73, 1187. Madsen, J., Hvelplund, T., 1985. Protein degradation in the rumen. Acta Agric. Scand. 25 (Suppl.), 103. Merchen, N.R., Bourquin, L.D., 1994. Processes of digestion and factors influencing digestion of forage-based diets by ruminants. In: Fahey, G.C., Jr. (Ed.), Forage quality, evaluation, and utilization. ASA, CSSA, SSSA, Madison, WI, USA, pp. 564–612. Merry, R.J., Mc Allan, R.B., 1983. A comparison of the chemical composition of mixed bacteria harvested from the liquid and solid fractions of rumen digesta. Br. J. Nutr. 50, 701–709. Merz, W., 1968. Automatische Schnellmethode zur Stickstoffbestimmung. Fresenius Zeitschrift fuer Analytische Chemie, p. 237. Michalet-Doreau, B., Ould-Bah, M.Y., 1989. Estimation of the extent of bacterial contamination in bag residues and its influence on in sacco measurements of forage nitrogen degradation in rumen. In: Proceedings of the XVIth International Grassland Congress. Nice, France, pp 909–910. Michalet-Doreau, B., Ould-Bah, M.Y., 1992. In vitro and in sacco methods for the estimation of dietary nitrogen degradability in the rumen: a review. Anim. Feed Sci. Technol. 40, 57–86. Michalet-Doreau, B., Vérité, R., Chapoutot, P., 1987. Méthodologie de mesure de la dégradabilité in sacco de l’azote des aliments dans le rumen. Bull. Techn. C.R.Z.V. (INRA, Theix) 69, 5–7. National Research Council, 1989. Nutrient Requirements of Dairy Cattle. 6th rev. ed. National Academic Press, Washington, DC. Ørskov, E.R., McDonald, I., 1979. The estimation of protein degradability in the rumen from incubation measurements weighted according the rate of passage. J. Agric. Sci. (Camb.) 92, 499–503. Ørskov, E.R., Macleod, N.A., Kyle, D.J., 1986. Flow of nitrogen from the rumen and abomasum in cattle and sheep given protein-free nutrients by intragastric infusion. Br. J. Nutr. 56, 241–248. Ould-Bah, M.Y., 1989. Adaptation de la technique in sacco à l’étude de la dégradation dans le rumen de l’azote des fourrages et application à l’étude des fourrages verts et conservés, thèse de doctorat. Académie de Montpellier, Montpellier, France, 186 pp. Perez, J.F., Balcells, J., Guada, J.A., Castrillo, C., 1996. Determination of rumen microbial nitrogen production in sheep: a comparison of urinary purine excretion with methods using 15 N and purine bases as markers of microbial-nitrogen entering the duodenum. Br. J. Nutr. 75, 699–709. Rémond, D., Le Guen, M.P., Poncet, C., 2003. Degradation in the rumen and nutritional value of lupin (Lupinus albus L.) seed proteins effect of extrusion. Anim. Feed Sci. Technol. 105, 55–70. Robinson, P.H., Fadel, J.G., Tamminga, S., 1986. Evaluation of mathematical models to describe neutral detergent residue in terms of its susceptibility to degradation in the rumen. Anim. Feed Sci. Technol. 15, 249–271. Šebek, L.B.J., Everts, H., 1999. In situ rumen degradation of dry matter and crude protein in ewes and dairy cows. Anim. Sci. 68, 801–808. Siddons, R.C., Evans, R.T., Beever, D.E., 1979. The effect of formaldehyde treatment before ensiling on the digestion of wilted grass silage by sheep. Br. J. Nutr. 42, 535–545. Sniffen, C.J., O’Connor, J.D., Van Soest, P.J., Fox, D.G., Russell, J.B., 1992. A net carbohydrate and protein system for evaluating cattle diets. II. Carbohydrate and protein availability. J. Anim. Sci. 70, 3562–3577. Tamminga, S., Chen, X.B., 2000. Animal-based techniques for the estimation of protein value of forages. In: Givens, D.I., Owen, E., Axford, R.F.E., Omed, H.M. (Eds.), Forage Evaluation in Ruminant Nutrition. CABI Publishing, Wallingford, UK, pp. 215–231. Tamminga, S., Van Straalen, W.M., Subnel, A.P.J., Meijer, R.G.M., Steg, A., Wever, C.J.G., Blok, M.C., 1994. The dutch protein evalution system: the DVE/OEB-system. Livest. Prod. Sci. 40, 139–155. Thomas, J.W., Yu, Y., Middleton, T., Stallings, C., 1982. Estimations of protein damage. In: F.N. Owens (Ed.), Protein Requirements of Cattle: Symposium. Oklahoma State University Press, Stillwater, pp. 81–98. Van Amburgh, M.E., Fox, D.G., Galton, D.M., Bauman, D.E., Chase, L.E., 1998. Evaluation of National Research Council and CNCPS net carbohydrate and protein systems for predicting requirements of Holstein heifers. J. Dairy Sci. 81, 509–526.

J.M.J. Gosselink et al. / Animal Feed Science and Technology 116 (2004) 35–51

51

Van Soest, P.J., 1982. Nutritional Ecology of the Ruminant. CNCPS University Press, New York, USA. Van Soest, P.J., Robertson, J.B., Lewis, B.A., 1991. Methods for dietary fiber, neutral detergent fiber, and nonstarch polysaccharides in relation to animal nutrition. J. Dairy Sci. 74, 3583–3597. Vanzant, E.S., Cochran, R.C., Titgemeyer, E.C., Stafford, S.D., Olson, K.C., Johnson, D.E., St Jean, G., 1996. In vivo and in situ measurements of forage protein degradation in beef cattle. J. Anim. Sci. 74, 2773–2784. Vérité, R., Michalet-Doreau, B., Chapoutot, P., Peyraud, J.L., Poncet, C., 1987. Révision du système des protéines digestibles dans l’intestin (P.D.I.). Bull. Tech. C.R.Z.V. (INRA, Theix) 70, 19–34. Yang, J.H., Broderick, G.A., Koegel, R.G., 1993. Effect of heat treating Alfalfa hay on chemical composition and ruminal in vitro protein degradation. J. Dairy Sci. 76, 154–164.